diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-05-23 04:56:47 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-05-23 04:56:47 -0500 |
| commit | 152821462023690df5d2bf90812e1cb5b1ca7274 (patch) | |
| tree | 9359c27d81b41dc0372cb82ef9c0ec3540d254e3 /scripts/train_hrm_orth.py | |
| parent | d11a0f6432e26c3243123d5e19aaf2702c76d64c (diff) | |
Add SRM training pipeline
- config/arch/srm_v1.yaml: arch config for pretrain.py integration
- scripts/train_srm.py: standalone from-scratch trainer based on step4
(HRM training infra adapted for SRM joint operator)
The arch.yaml exposes κ, η, α, n_iters, n_aol_layers as Hydra params.
train_srm.py adds joint Lyapunov diagnostic via JVP on srm_block to verify
λ_1 ≤ log((1-α)+α·κ) per micro-step. Smoke tested with hidden=128, n_iters=4
on Sudoku 1k: empirical Lip 0.28 << bound 0.90.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Diffstat (limited to 'scripts/train_hrm_orth.py')
0 files changed, 0 insertions, 0 deletions